NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-06-01KK1GWSHWY7ET8GFWQECBVFE8
Run Details
Model:
GPT-5.4
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
10.5s
Cost:
$0.005160
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
93.65%
Matches word count
(details)
36.79%
Dialogue to Total Word Ratio
(details)
65.2188%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 90% words of dialogue between two thieves named Molly and Bill who are planning to break into a children's party to steal all the presents. Only include the prose of the scene without any markdown or commentary.
Model response
Messages
Meta
AI
“Bill, you cannot wear the eye patch.” “Why not? It says menace.” “It says you lost a fight with a stapler.” “The patch stays. Now, presents. We go in smiling, leave with loot.” “At a children’s party?” “Children are basically very short drunks. Unpredictable, sticky, easily distracted.” “Your research alarms me.” “Focus, Molly. Balloons at noon, cake at one, screaming all day. We slip in as entertainers.” “You can’t juggle.” “I can drop things rhythmically.” “And I’m what, the clown?” “No, you’re the face painter. You already own suspiciously many brushes.” “I paint watercolors, Bill, not tigers on feral six-year-olds.” “Same wrists, different canvas.” “What about the parents?” “Helpless. One’s inflating balloons, one’s arguing about gluten, three are pretending juice isn’t mostly sugar.” “And the presents?” “Stacked by the hallway table. We sweep them into the laundry hamper.” “That’s your grand criminal vessel? A hamper?” “Nobody questions a hamper. It whispers chores.” “What if a child sees us?” “We hand them a whistle.” “That creates more witnesses.” “Louder witnesses, yes, but less coherent.” “Bill, this is appalling.” “Efficient.” “It’s monstrous.” “It’s gift redistribution.” “It’s stealing.” “It’s seasonal.” “It’s June.” “Unseasonal, then.” “Fine. But if anyone asks, we’re magicians.” “Can you do magic?” “Absolutely. Watch me make a bicycle disappear.”
created_at
"2026-03-06T12:10:26.755Z"
3/6/2026, 12:10:26 PM
model
"openai/gpt-5.4-20260305"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
417
moderation_latency
(null)
generation_time
10170
tokens_prompt
77
tokens_completion
434
native_tokens_prompt
66
native_tokens_completion
333
native_tokens_completion_images
0
native_tokens_reasoning
0
native_tokens_cached
0
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"completed"
usage
0.00516
router
(null)
provider_responses
0
endpoint_id
"9ff5625c-403f-4d7f-b895-58ac7295062c"
id
"resp_08051f99920476c00169aac432ce5c8193b5837d033141502b"
is_byok
false
latency
142
model_permaslug
"openai/gpt-5.4-20260305"
provider_name
"OpenAI"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.0; linux; x64))"
http_referer
(null)
api_type
"completions"
id
"gen-1772799026-tnR8uv5YQjam17LjuvBq"
upstream_id
"resp_08051f99920476c00169aac432ce5c8193b5837d033141502b"
total_cost
0.00516
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
93.65%
Matches word count
n/a
neededClean
false
words
209
36.79%
Dialogue to Total Word Ratio
Ratio: 100.00%, Deviation: 10.00%
neededClean
false
wordsTotal
211
wordsDialogue
211
65.2188%